Human face super-resolution algorithm based on regional depth convolution neural network

Abstract

The invention discloses a human face super-resolution algorithm based on regional depth convolution neural network. The algorithm comprises the following steps: a training stage: S1) dividing the mutually overlapping image blocks in the pixel domain of an inputted human face image with low resolution to obtain a plurality of local regions; S2) extracting the local regions for local characteristics; S3) performing non-linear change to the local characteristics to obtain non-linear characteristics; S4) processing the non-linear characteristics to obtain reconstructed image blocks with high resolution; S5) splicing the image blocks with high resolution; adjusting the multi-layer convolution layers and correcting the parameters of the linear unit layer; and a testing stage: S6) inputting the tested human face image with low resolution; processing through the super-resolution network to obtain the human face image with high resolution. The regional convolution neural network proposed by the invention improves the quality of subjective and objective reconstruction of reconstructing high resolution images.

Description

technical field [0001] The invention relates to the field of image super-resolution, in particular to a face super-resolution algorithm based on a regional deep convolutional neural network. Background technique [0002] Simon and Kanade et al. proposed a face hallucination method specifically for face images. Face hallucination is an image super-resolution that generates high-resolution face images from input low-resolution face images. rebuild method. [0003] Traditional reconstruction-based super-resolution algorithms are suitable for small magnifications. When the magnification increases, the reconstruction method cannot obtain more prior knowledge, which limits its reconstruction quality. The face super-resolution algorithm based on learning can be divided into single-layer method and multi-layer method. The single-layer method uses the local blocks of the face image to share the maximum similarity and obtain the optimal weight vector. The method is to use the deep ne...

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is to provide a face hypersensing method based on a regional deep convolutional neural network in view of the lack of expressive ability of the single-layer learning structure in the prior art and the inability to make full use of the structural knowledge on the face image. resolution algorithm

Images